MicroRNAs in rheumatoid arthritis: what is the latest with regards to diagnostics?

KEYWORDS:

• MicroRNAs
• biomarker
• rheumatoid arthritis
• diagnostics

Rheumatoid arthritis (RA) is a systemic, chronic and inflammatory disease with unknown etiology [1]. Characteristic features of RA are chronic synovitis and a fluctuating clinical course, eventually leading to joint deformities or even long-term disability [1], and thereby carrying a substantial burden for both the individual and society [2]. Early diagnosis is the key for successful treatment, particularly in patients with risk factors for poor prognostic (including high disease activity, presence of autoantibodies, and early joint damage) [3]. Although standard diagnostic criteria are now available for existed RA diagnosis, the most important challenging issue remains to find biomarkers for early diagnosis, especially for patients who cannot be made a specific diagnosis at first presentation, and their exclusive diagnosis is undifferentiated arthritis [4]. Accumulating evidences show that microRNAs (miRNAs) can be aberrantly expressed in both inflamed synovium and circulation of RA patients [5], and the role of miRNAs in RA might fulfill the criteria for their using as novel molecular diagnostic markers for RA [6]. Here, we read with great interest the articles and provide an expert review of miRNAs in diagnostics in patients with RA.

miRNAs are endogenous, single-stranded, non-coding RNAs that are evolutionarily conserved with a length of 18–25 nucleotides [7]. miRNAs can influence the target mRNA processing at the post-transcriptional level by promoting mRNA degradation or translational repression [8]. The first evidence for miRNAs’ role in human diseases was shown in cancer cells, which opened up the exciting prospect of using miRNAs as powerful, non-invasive biomarkers for immune responses and autoimmune diseases [9]. In 2007, Bhanji and colleagues first discovered that miRNAs might play a role in RA progression [10]. Over the following year, subsequent studies reported the dysregulation of certain miRNAs within inflamed joints and in the peripheral circulation of patients with RA, and these specific miRNAs were miR-16, miR-132, miR-146a, and miR-155 [11–13]. Since then, much has been learned about the majority of miRNAs involved in the natural course of RA, and those studies to date have mainly focused on synovial tissue/fluid and peripheral blood from established RA, often in minimally defined and heterogeneous cohorts.

The dysregulation of miRNAs in inflamed joints from RA has been demonstrated by a number of works in literature. There is a study showing that the expression levels of miR-16, miR-146a, miR-155 and miR-223 were significantly increased in synovial fluid of RA patients, which may differentiate RA from osteoarthritis (OA). Several other studies showed that in RA synovial fibroblasts (RASFs) isolated from inflamed joints of RA patients, miR-133a, miR-142-3p, miR-142-5p, miR-146a, miR-155, miR-203, miR-221/miR-222 cluster, miR-223 and miR-323-3p presented with the increased expression, while the expression of miR-124a and miR-34a* were decreased [13–18]. Mechanism study showed that both miR-124a and miR-34a* can inhibit synovial cell proliferation [15,16]. In synovial tissue, miR-146a, miR-146b, miR-150, miR-155, miR-223 were obviously high-expressed, while miR-22, miR-23b, and miR-30a were downregulated [12,13,19–23]. Moreover, in synovial fluid CD4+cell in RA joints, the reports showed that the miR-21, miR-363 and miR-498 were significantly low-expressed, while the expression level of miR-146a was changed in an opposite trend [24]. In addition, there was an upregulated expression of miR-146 and miR-155 in both synovial fluid CD4+cell and synovium macrophages [11–13,25].

It is believed that synovial tissues, RASFs, and peripheral blood mononuclear cell (PBMC) secret miRNAs in distinct patterns [14]. In PBMC, seven miRNAs (miR-16, miR-26a, miR-132, miR-146a, miR-146b, miR-150, and miR-155) were reported to be significantly increased, and interestingly only miR-21 was found to be significantly decreased [11,22,26].

The abnormal expression of miRNAs in the blood of patients with RA was also discovered and addressed in many studies. There are amount data showing that the expression levels of nine miRNAs (miR-16, miR-21, miR-24, miR-26a, miR-125a-5p, miR-125b, miR-126-3p, miR-223 and miR-451) were increased in serum/plasma from RA patients whereas six miRNAs (miR-16, miR-125a-3p, miR-126-3p, miR-132, miR-146a and miR-155) were significantly downregulated [14,27–30]. Of note, the miR-16 and miR-126-3p can be upregulated and also downregulated in RA serum/plasma, which is controversial and need to be re-studied. In the whole blood, many miRNAs were found to be significantly overexpressed, and they are miR-99, miR-100, miR-125b, miR-146a and miR-155 [29,31]. Finally, there is a study reporting that among the peripheral blood miRNAs, nine miRNAs (miR-16, miR-21, miR-24, miR-26a, miR-125a-5p, miR-125b, miR-126-3p, miR-223, and miR-451) could be the most promising biomarkers for the detection of RA [32].

Importantly, plasma miRNAs were found to be correlated with the disease activity of RA. For example, it is reported that plasma miR-16 was correlated with RA disease activity indexes, such as tender joint count (TJC) and 28-joint Disease Activity Score (DAS28) [14]. Moreover, circulating miRNAs can be predictive markers for treatment response. For instance, the high serum level of miR-125b at disease flare was associated with good clinical response to treatment with rituximab three months later, and this predictive value was not limited to RA as it was also found in B lymphomas patients [29]. These raise the possibility of diagnostic values of plasma/serum miRNAs. Intriguingly, plasma miR-24 and miR-125a-5p were found to be the potential diagnostic markers of RA even if patients were anti-citrullinated protein antibody (ACPA)-negative [28].

In addition to established RA, miRNAs in early rheumatoid arthritis (ERA) also have driven Rheumatologists’ attention. MiR-146a, miR-155, and miR-16 were found to be decreased in the serum of ERA patients, as compared to established RA [30]. Moreover, in this study, a change in circulating miR-16 in the first 3 months of therapy was associated with a decrease in DAS28 in long term follow-up in ERA, and circulating miR-223 level in treatment naïve ERA was correlated with C reactive protein (CRP), DAS28 and the change in DAS28 after 3 months and 12 months of follow-up [30]. Therefore, the authors concluded that miR-223 is a marker of disease activity and both miR-16 and miR-223 are possible predictors for disease outcome in ERA [30]. Recently, another work in literature from Ouboussad et al. firstly showed that serum miR-22 has a potential biomarker role for informing a predictive of progression from ‘at-risk’ to RA by comparing miRNA expression among healthy individuals, those at risk of and those that develop RA [33]. It is the first miRNA study focused on individuals at risk of RA, which suggested that miRNAs might have a predictive role in the progression from ‘at-risk’ status to RA.

Immune cells play a key role in pathogenesis of RA. Recently, a study defined miRNAs that are specifically expressed in both naive and memory regulatory T cells (Tregs), and such miRNA signature characterized the Treg phenotype (miR-146a, miR-3162, miR-1202, miR-1246 and miR-4281) [34]. miR-34b, miR-223, and miR-451 were reported to have an up-regulation in peripheral blood T cells and miR-146a was increased in peripheral blood CD4+cells [27,34,35], while there were two miRNAs (miR-21 and miR-363) and one miRNA (miR-451) downregulated in neutrophils and peripheral blood CD4+cells, respectively, [24,26,36]. These miRNAs expressed in immune cells are also associated with disease activity of RA. For example, it is reported that miR-146a expressed intensely in the synovium with hyperplasia and high expression of interleukin (IL)-17 from the patients with high disease activity [22]. In addition, the expression level of miR-223 in T cells was found to be correlated positively with the titer of rheumatoid factor (RF) in RA patients [35]. Moreover, increased expression of miR-451 in T cells of RA patients was positively correlated with DAS28, erythrocyte sedimentation rate (ESR) levels and serum levels of IL-6 [34].

Collectively, a lot of miRNAs have been found to be aberrantly expressed in either synovial tissue/fluid, blood, or immune cells of RA patients (the information and function of several key miRNAs are summarized in Table 1), though sometimes the data are controversial. Therefore, the combinations of different miRNAs may increase diagnostic accuracy, particularly for further studies in the undifferentiated arthritis and ACPA-negative RA.

Table 1. Information of several well-documented miRNAs in RA.

MiRNAs	Expression	Cells/organ of RA	Roles/functions	Reference
miR-146a	↑	RASFs treated with TNF-alpha Synovial fluid PBMC IL-17 producing T cells CD4+T cells of RA patients	Regulation of TNF-alpha production Associated with TJC and DAS28 Expressed intensely in the synovium with hyperplasia and high expression of IL-17 from the patients with high disease activity. suppresses T cell apoptosis	[11,13,14,22,24,31]
miR-155	↑	RASFs treated with TNF-alpha PBMC Synovial tissue Synovial fluid Synovial fluid monocytes synovial membrane and synovial fluid macrophages synovial CD14+cells	Downregulation of SHIP-1 Represses the levels of MMP-3 MiR-155-deficient mice are resistant to collagen-induced arthritis, with a profound suppression of antigen-specific Th17 cell and autoantibody responses and markedly reduced articular inflammation.	[11,13,14,22,25,31]
miR-223	↑	RA synovium: the superficial and sublining layers, and macrophages, monocytes, and CD4 T cells Synovial fluid Serum Circulating T cells	Suppresses osteoclastogenesis in vitro A marker of RA disease activity A possible predictor for disease outcome in ERA Correlated positively with the titer of RF	[14,23,30,35]
miR-203	↑	RASFs	Promotes MMP-1 and IL-6 production	[18]
miR-132	↓	Plasma	Plasma miR-132 can be a potential diagnostic marker for patients with RA and OA.	[14]
miR-34a*	↓	RASFs	Lead to an increased rate of FasL- and TRAIL-mediated apoptosis in RASFs Highly correlated with expression of XIAP, which was found to be upregulated in RA synovial cells.	[15]
miR-124a	↓	RASFs	A key miRNA in the post-transcriptional regulatory mechanisms of RASFs	[16]
miR-23b	↓	RASFs	Represses autoimmune inflammation	[19]
miR-30a	↓	Synovial tissue	Suppresses autophagy	[20]
miR-451	↓	Circulating T cells Neutrophils	A positive correlation between increased expression of miR-451 in T cells of RA patients and DAS28, ESR levels and serum levels of interleukin-6 Suppresses neutrophil chemotaxis	[34,36]
miR-22	↓ ↑	Synovial tissue Serum	miR-22 was downregulated in RA synovial tissue, contributing to the overexpression of Cyr61 in RA FLS A potential biomarker role for individuals at risk of RA	[21,33]

Abbreviations: DAS28, 28-joint Disease Activity Score; ERA, early rheumatoid arthritis; ESR, erythrocyte sedimentation rate; IL, interleukin; MMP, matrix metalloproteinase; RA, Rheumatoid arthritis; RASFs, rheumatoid synovial fibroblasts; RF, rheumatoid factor; SHIP-1, SH2-containing inositol phosphatase-1; TJC, tender joint counts; TNF, tumor necrosis factor; PBMC, Peripheral blood-derived mononuclear cells; XIAP, X-linked inhibitor of apoptosis protein.

The favorable biological characteristics of miRNAs are taking into account the important factors for miRNAs to serve as a diagnostic biomarker. Firstly, miRNAs can be easily measured and separated from various sources, including blood, biological fluids, tissue sample, and even formalin-fixed paraffin-embedded samples [37]. Secondly, miRNAs have high stability in plasma or serum as well as under long-term storage conditions [14], attributing to the circulation in membrane vesicles which protected the miRNAs from degradation by endogenous RNases [38].

However, in spite of several advantages of miRNAs as potential biomarkers for RA, there are some challenges of miRNAs application for diagnostic purposes. First of all, although a list of available evidences revealed that the dysregulated expression of a dozen miRNAs in patients with RA is different from those expressed in OA or healthy people, none of the miRNAs identified in RA tissues are specific for RA. Koichi Murata and coworkers have investigated the concentrations of five miRNAs (miR-16, miR-132, miR-146a, miR-155, and miR-223) in plasma and synovial fluid, as these miRNAs were reported to be potential diagnostic markers for RA. Unfortunately, only miR-132 can differentiate patients with RA from healthy control subjects in low plasma levels, and plasma miR-132 levels cannot discriminate between RA and OA [14]. It means that none of the five miRNAs quantified is able to differentiate between RA and OA. What is worse, the in-depth study performed by Murata et al. detected that although the expression pattern of miRNAs in the synovial fluid was similar with those secreted by the synovial tissue, the miRNAs in synovial fluid and plasma have different origins, and there is no direct correlation between plasma miRNA concentrations and synovial fluid miRNA levels. This is in line with reports showing that miR-146a was increased in RASFs [13,16–18], synovial tissue [12,13,19,22], synovial fluid CD4+cells [24], and PBMCs [11,22], while decreased in serum/plasma [27,30]. Thus, more studies of additional miRNAs in order to obtain homogeneous expression in synovial and serum samples are required. Secondly, the technical difficulties of examining serum samples were concerned as another challenge, because of low miRNA abundance, inaccurate miRNA quantification assays, and the lack of accepted and standardized methods for normalization [39]. Fortunately, extraction of exosomes from serum and using the mixed approach (i.e. the combination of high-throughput experimental procedures coupled with computational methods) to decipher and validate new miRNAs are quite promising [40,41].

Last, we appreciated the researchers for their constructive and meaningful works on the role of miRNAs in RA, which gives perspectives on the use of this miRNA as the novel molecular diagnostic markers for RA patients and agreed with the importance of specific miRNAs for future studies. We hope that miRNAs can be considered as potential diagnostic biomarkers for RA, although there seems to be a long way to go.

Declaration of interest

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Reviewer disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Additional information

Funding

This manuscript was funded by the grants from Guangdong Provincial Hospital of Chinese Medicine (No. YN10101906, No. YN2018ML08, and No. YN2018ZD06) and Guangdong Science and Technology Department (No. 2016A020226041).